Nonsurgical Management of Type II Fractures

Nonsurgical treatment of type II odontoid fractures consists of either rigid collar fixation or halo immobilization. Halo-vest treatment has been shown to limit upper cervical motion by 75% versus 45% with a collar.³¹ The drawbacks with halo immobilization are many, including economic, social, pin site infections, skin breakdown, skull perforation, and overall nonunion on the order of about 32%.^{2,6,24,26,32–40}

Several studies have attempted to characterize the risk factors involved in nonunion of type II fractures. Platzer and colleagues found that 75 of 90 patients returned to their preinjury activity level and were satisfied with halo–thoracic vest treatment. Successful fracture healing was achieved in 84% of patients undergoing this treatment. Cases in which nonunion was demonstrable on radiographs or CT scans within 6 to 12 months after trauma were attributed to increased age or greater dens displacement.²⁴

Seybold and Bayley found a 65% overall fusion rate with halo immobilization and no significant difference in outcome in patients older than 60 years, in patients younger than 60 years, or by degree of displacement.⁴¹ Greene and colleagues reviewed 340 acute axis fractures and identified two categories of unstable type II fractures, which they recommended be treated by early surgical fusion. The first involves dens displacement of 6 mm or greater, and the second involves comminution of the dens.⁴² Hadley and coworkers found a 67% nonunion rate in patients with 6-mm or greater subluxation as opposed to a 9% nonunion rate in those with lesser degrees of displacement. They failed to find any correlation between a patient’s age or direction of subluxation.^43–45 Dunn and Seljeskog found an increased risk for nonunion with conservative management in patients older than 65 years and in those with posteriorly displaced odontoid fragments. They reported a nonunion rate of 70% with retrolisthesis versus 30% with anterolisthesis.⁴⁶

Because of these risk factors for nonunion (age >65 years, >6-mm displacement, posterior displacement), surgical management evolved as an appropriate treatment modality. Indications for surgery include irreducible or unstable fracture for which reduction is impossible; significant neurological deficit; previous nonunion; prolonged delay in treatment; injuries that preclude halo placement, including skull fractures and chest or facial injuries; and patients at high risk for nonunion, including the elderly and debilitated.^47–49

Surgical management of these injuries can be approached either anteriorly or posteriorly. Posterior approaches involve the use of wire or cable, clamps, or direct screw fixation.⁵⁰ These techniques involve C1-2 fusion and loss of motion at this joint. Loss of rotary motion at C1-2 is usually in the 50% range.^51,52 Anterior procedures involve direct screw fixation across the fracture line. Traynelis and associates performed an evidence-based review of posterior bone and wire fixation of odontoid fractures and noted an overall success rate of 64%, with the morbidity and mortality rate approaching 2%.^53,54 A variety of bone and wire fusion techniques have been described, including the Gallie (midline graft with a single wire), Brooks (bilateral sublaminar wires), and Sonntag (interspinous) methods. These techniques create a posterior tension band. Although successful, these techniques are not strong in extension and axial rotation. To overcome these deficiencies, Magerl described the technique of transarticular C1-2 fixation.^76–78 This method involves screw fixation through the lateral joints of C1-2. Extensive preoperative planning is required to determine the course of the vertebral arteries because they may be injured by the trajectory of the screw. Harms and Melcher recently described a construct consisting of direct lateral mass screw fixation of C1 and pars or pedicle screw fixation of C2.⁵⁸ This technique has proved to be biomechanically sound and technically easier than transarticular fixation, does not require the use of midline bone grafts, and can be performed despite C1 laminectomy.

Because of the potential morbidity and loss of motion with posterior C1-2 fusion, the use of anterior screw fixation for type II odontoid fractures has become increasingly popular. This technique provides immediate stabilization; we have performed this procedure and often discharged the patient the next day. The first reports of odontoid fixation were reported by Nakanishi in 1980⁵⁹ and Bohler in 1982.⁶⁰ These early reports involved extensive neck dissection and were not widely adopted.

As surgical techniques became more refined, direct anterior screw fixation of C2 fractures has become more popular, and many clinicians now consider this procedure to be the treatment of choice for these injuries. Fusion rates in several series have ranged from 79% to 100%, with acceptably low morbidity.^{25,27,61–64} Apfelbaum and coworkers reviewed their series of 147 patients who underwent anterior screw fixation and found an overall fusion rate of 77%, which rose to 88% with acute fractures.⁶⁵ An analysis of patient factors noted that fracture orientation was the only significant variable that affected fusion rates. Fractures with anterior oblique orientation were more likely to be associated with fusion failure or fibrous union than were posterior oblique or horizontal fractures. The two most common hardware failures were screw backout and screw pullout. Screws that backed out were associated with lack of engagement of the distal cortex of the odontoid tip. Screws pulled out of the body of C2 in patients who had fractures that involved the C2 body. Current recommendations are posterior stabilization for severely comminuted fractures of the body of C2. Platzer and coauthors recently reported a series of 79 patients who underwent anterior screw fixation and found that patients younger than 65 years had better outcomes than those older than 65.²⁵

Indications for Anterior Odontoid Screw Fixation

As noted earlier, most patients with acute type II fractures are candidates for anterior screw fixation. This approach allows direct fixation across the fracture line and provides immediate stabilization of the injury while maintaining motion at the C1-2 joint. Contraindications to this technique include rupture of the transverse ligament (which can be assessed with MRI) and significant comminution of the C2 body. Relative contraindications include osteopenia, which affects pullout strength of the screw, and anterior obliquely sloping fractures, which can allow the odontoid process to slide along the fracture line as the screw pulls it inferiorly. This tends to be less of an issue with acute fractures but is more important with fractures more than several weeks old. Apfelbaum and coworkers noted that anterior screw fixation can generally be performed for fractures up to 6 months old; in patients with chronic nonunion, scar tissue develops and makes screw passage difficult.⁶⁵ Patients with contraindications to anterior screw fixation are optimally treated with transarticular screw fixation or the Harms C1-2 technique. Patients with type III odontoid fractures and ruptured transverse ligaments are also best treated with a posterior procedure; if the ligament is intact but there is motion at the fracture site, anterior screw fixation can be considered.

Anterior Odontoid Screw Fixation

Several authors have described anterior C2 osteosynthesis in detail; for a more detailed description of the procedure, the reader is referred to other sources.^{24,25,63,65–68}

If the fracture is unstable, awake fiberoptic intubation is necessary. The patient is placed supine on the operating table with a shoulder roll to facilitate neck extension. Tong traction is also used. Anteroposterior (AP) and lateral fluoroscopy is necessary to ensure reduction of the fracture and accurate screw fixation. A lateral image is obtained to assess reduction of the fracture. If reduction is achieved, the current alignment is maintained. If the malalignment continues, the head is placed in neutral position. The patient is then prepared and draped in the usual sterile manner.

A standard transverse skin incision is made over the C5-6 disk space. The platysma muscle is opened, and blunt dissection is used to access the corridor to the ventral spine. We often approach the cervical spine from the left side to avoid injury to the recurrent laryngeal nerve, which has a more variable course on the right. The trachea and esophagus are medial and the carotid artery lateral to the dissection. Once the ventral spine is palpated, handheld retractors are used, and the longus colli is incised (often with cautery) and swept laterally off the midline. Self-retaining retractors are now placed in the incised muscle. The soft tissue is opened cephalad to the C2 region to allow access to the C2-3 disk space.

At this point in the procedure, cannulation of the fracture can begin. Two similar techniques have been described—the cannulated screw technique and the standard lag screw technique. In the cannulated screw technique, the C2-3 anulus is opened as the starting point for drilling, and a Kirschner wire (K wire) is inserted at the anterior lip of C2 in a sagittal direction such that it is angled posteriorly and exits the posterior tip of the odontoid. AP fluoroscopy is useful to ensure that the K wire is in the midline, and a lateral view ensures the appropriate trajectory. A drill guide system is placed over the K wire once the pilot hole has been made. The drill is now passed through the body of C2 to the fracture line, and after the surgeon is assured that the spine remains aligned, a hole is drilled to and then through the posterior apex of the odontoid. The drill is removed, the pilot hole is tapped, and a lag screw is inserted through the guide tube through the fracture. It should be emphasized that frequent imaging is useful in achieving optimal trajectory through the fracture. In the lag screw technique, a drill bit is inserted into the anterior edge of C2, and a hole is drilled without the use of a K wire. A screw is then inserted through the drill guide. A hand drill or power drill can be used. Biomechanical studies have shown that one screw is as stable as two screws.^{7,25,65,69,70–72}

The wound is closed in standard fashion. We often place the patient in a collar for 4 to 6 weeks. The patient is usually discharged within 48 hours. Several studies have documented the high rate of healing and low morbidity of this operation^65,72–75; it is possible that direct anterior screw fixation of odontoid fractures will become the treatment of choice in young patients with acute type II fractures (Fig. 315-3).

FIGURE 315-3 A-E, Forty-four-year-old man seen after a fall. He was neurologically normal but complained of neck pain. Radiographs showed a type II odontoid fracture. The patient underwent anterior odontoid screw fixation and was discharged the next day.

Transarticular Screw Placement for C2 Fractures

Transarticular screw fixation of C1-2, originally described by Magerl and Seemann,⁷⁶ is a useful surgical technique in patients who are not candidates for anterior C2 screw fixation.^17,77–81 Such patients include those with poor bone stock, such as the elderly (Fig. 315-4); patients whose fractures are older than 6 months; and patients with an injured transverse ligament. The procedure is technically more challenging than anterior C2 surgery and necessitates an intact C1 arch because interspinous wiring is also required to ensure long-term stabilization.^67,81

FIGURE 315-4 A-C, Sixty-two-year-old white man with a type II odontoid fracture. Reformatted computed tomography scans show ample space for screw trajectory across the C1-2 joint.

Careful preoperative preparation is needed for this technique. Thin-cut CT scans in multiple planes are obtained to assess the fracture pattern, as well as determine the course of the vertebral artery. If the vertebral artery traverses aberrantly through a path that crosses a planned screw trajectory, a screw should not be attempted on this side.

If preoperative instability exists, awake fiberoptic intubation should be performed. Neck movement during intubation can be assessed via fluoroscopy because this will be required for surgery. The patient is placed prone in a head holder after induction of anesthesia, and biplanar fluoroscopy is used to assess spinal alignment. The C-arm can be draped into the field once surgery has commenced. Flexion of the neck, as much as safely possible, facilitates screw insertion.

A midline incision to C5 is used to expose the lateral aspects of the C2-3 facet joints, as well as the C1 arch. Hemostasis can be achieved with monopolar and bipolar cautery, and self-retaining retractors are used to maintain visualization of the field. Care must be taken to limit lateral dissection of C1 to prevent injury to the vertebral artery.

Once C2-3 exposure is completed, attention is turned to the entry point of the screw at C2. A dissector or curet is used to expose the cephalad laminar surface and isthmus (pars) of C2. The atlantoaxial membrane is visible medial to the pars. The pars should be exposed upward to the C1-2 joint and the C2 nerve root identified. The joint can easily be probed with a dissector. Bleeding in this area, which can be brisk, can be controlled with bipolar cautery, FloSeal, or thrombin-soaked Gelfoam.

Once both pars have been exposed, the entry point for drilling can be identified. It is at the lower edge of the caudal articulating process of C2, 2 mm lateral to the medial edge of the pars, and about 3 mm cranial to the C2-3 joint. Depending on which system is used, an obturator may need to be inserted through a separate stab incision, or a flexible drill head can be used. A K wire can be used to get an idea of screw length and the required trajectory through the C1-2 joint.

A drill is now placed through the drill guide, and with the alternating use of AP and lateral fluoroscopy, a drill hole of appropriate length and trajectory is made. The path of the drill hole goes through the C1-2 joint and enters the lateral mass of C1, pointed at the anterior tubercle. Bicortical purchase is desirable. The preoperative CT scan should be available to ensure correct trajectory in three dimensions.

Once the drill hole is completed, the drill guide on the drill or a depth gauge allows determination of appropriate screw length. The screw hole is tapped, and then a fully threaded screw is placed through the hole. The contralateral screw is next placed similarly.

If arterial bleeding is noted from the first screw hole, a screw can be inserted to control the bleeding. The contralateral screw should not be placed, however, because of the potential for brainstem stroke from bilateral vertebral artery injury. A cerebrovascular accident from unilateral injury is rare.

Once both transarticular screws have been placed, C1-2 interspinous wiring and bony fusion should be performed (Fig. 315-5). This can be done with the Brooks or Sonntag method and increases the stability of the construct so that bony fusion can take place. Bone graft can also be placed in the lateral gutters.⁸²

FIGURE 315-5 A-C, Same patient as in Figure 315-4 after C1-2 fusion with transarticular screw fixation and posterior wiring with a graft.

Transarticular screw fixation has high rates of fusion but is technically challenging. It requires biplanar fluoroscopy, as well as C1-2 wiring. For this reason, several surgeons have turned to direct C1-2 screw fixation (the Harms technique) for the management of complex C1-2 instability.

Posterior C1-2 Screw-Rod Fixation

Despite its high fusion rate and biomechanical superiority to wiring techniques, there are some drawbacks to C1-2 transarticular fixation, including technical challenges, the need for an intact C1 arch, screw breakage, the need for C1-2 wiring and a structural graft, difficult reduction of the C1-2 deformity, and possible catastrophic injury to the vertebral artery.

Posterior C1-2 screw-rod fixation overcomes some of these limitations. First described by Harms and Melcher,⁵⁸ this procedure involves bilateral placement of screws with polyaxial heads into the C2 pedicles and lateral masses of C1 (Fig. 315-6). Fluoroscopy may be necessary to assess reduction of the subluxation and final screw placement.

FIGURE 315-6 A-E, Fifty-eight-year-old woman fell and suffered an odontoid fracture. She underwent successful C1-C2 fixation. B, Plain x-rays; C-D, Axial CT scan; E, Reformatted sagittal CT scan.

Awake fiberoptic intubation may be required if the spine is unstable. Once anesthesia is induced, the patient is placed prone in a Mayfield head holder. Lateral radiography or fluoroscopy confirms alignment of the spine. Standard midline subperiosteal dissection of C1-3 is performed. C1 and C2 are exposed up to the lateral aspect of the C1-2 articulation. Careful dissection techniques allow avoidance of the vertebral artery as it travels laterally at C2 and then medially at C1. Venous bleeding, which can occasionally be vigorous, can be controlled with Gelfoam and cottonoids. Hemostasis can be achieved on one side while dissection continues on the contralateral side.

Once C1 and C2 have been adequately exposed, the dorsal root ganglion of C2 can be retracted downward. This maneuver exposes the entry point of C1, which is the middle of the lateral mass, halfway between the arch of C1 and the C1-2 joint. Occasionally, the arch of C1 may need to be drilled off or removed with a Kerrison punch to expose the entry point. A high-speed drill is used to break through the cortex, and then a hole is drilled with a drill and drill guide. The direction is slightly medial and parallel to the C1 arch.⁸³ Lateral fluoroscopy is useful for defining the trajectory; the ultimate target is the anterior cortex of the C1 lateral mass. An appropriate-length screw is placed bicortically once the hole is tapped. The screw is only partially threaded because there will be some length of the screw not engaged in C1.

The C2 screws can now be placed. The entry point is halfway between the upper and lower articular surfaces of C2. A high-speed drill can be used to make a pilot hole, and the hand drill and drill guide are oriented 25 degrees cephalad and 20 to 25 degrees medially. The hole can be probed sequentially as the drill guide is removed and then reinserted. Screw length is usually 25 to 35 mm. Rods are now inserted in the screw heads and tightened with caps. Fusion can be performed at the C1-2 joint.

Harms and Melcher reported on their series of 37 patients treated over a 3-year period.⁵⁸ Half of these patients were operated on for trauma, and no neural or vascular complications were reported.

This technique has gained wide acceptance in the spine surgery community and has been used frequently in elderly patients with type II odontoid fractures. This subgroup of patients may have preexisting comorbid conditions, and thus a straightforward, less technically demanding procedure may be particularly useful.